HBMNet, a hierarchical boundary-modeling network, outperforms BA-TFD and UMMAFormer at localizing partially manipulated audio-visual segments on a subset of AV-Deepfake-1M.
In 2022 IEEE Spoken Language Technology Workshop (SLT)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Localizing Audio-Visual Deepfakes via Hierarchical Boundary Modeling
HBMNet, a hierarchical boundary-modeling network, outperforms BA-TFD and UMMAFormer at localizing partially manipulated audio-visual segments on a subset of AV-Deepfake-1M.